Detecting Cultures

Ceramic red-slipped head lug.
YPM catalog no. 253686.


“Archaeology is like detective work.
The few cues we have determine where we start.” — Ben Rouse

Potsherds per se, Ben Rouse has written, are not parts of cultures; instead they are artifacts that peoples discarded because they were no longer useful. A consistent theme in Professor Rouse’s work is the development of methods of returning sherds to the cultures, now long gone, that produced them and the reconstruction of the prehistory of those cultures and the cultures around them in space—and time. Rouse’s work at the Indian Creek site on Antigua in the Caribbean provides a good example.

The West Indies.


By the time of that excavation in 1973, Professor Rouse and others had constructed a chronology for the Greater Antilles to the north of the Leeward Islands, and were about to construct another one for the Windward Islands to the south. Rouse envisaged the local sequence for Indian Creek as the first of several needed for the development of a chronology that would fill the gap between the other two. He hoped to test conflicting hypotheses on the migration of Ceramic Age peoples from South America into the West Indies.

As it turned out, he was able to go much further. In their study reported in Excavations at the Indian Creek Site, Antigua, West Indies (published by the Yale Peabody Museum), Rouse and co-author Birgit Faber Morse offer the first comprehensive regional chronology that can be used to test the migration hypotheses more conclusively.

Map of Antigua showing
the location of the Indian Creek site.


The settlers who followed Columbus paid little attention to the Lesser Antilles; Europeans did not colonize Antigua until 1632. Its first native agriculturalists had arrived late as well, about A.D. 1, perhaps because its low, variable and highly seasonal rainfall discouraged farming. The island’s area of greatest rainfall is in its volcanic region, at the southeastern corner of which Indian Creek sits.

The site’s name comes from a dried-up streambed that emptied into a nearby small rocky cove. Professor Rouse first surveyed the site in 1959, but was not able to work on it until more than a decade later when, overcommitted and overworked, he fell ill from heat exhaustion after only a month, and co-workers completed his plan.

Aerial photograph of the Indian Creek site, looking south.
Courtesy of the Fred Olsen Trust.


Indian Creek had many features that attracted archeological attention. The site had survived with little damage thanks to the circumstances of its discovery and its seclusion. The site was also, after years of drought, almost bare of vegetation. The many conch and other shells visible around the edge of the site were evidence of prehistoric occupation; specimens of petrified wood showed that trees once grew in this part of Antigua.

From the air, the site, which covers more than 20 acres, resembles an oval ring, light gray against a darker background. The gray comes from the concentration of shell refuse in 6 middens around the periphery. The first midden to be excavated, the smallest, shallowest and farthest from the richest concentration of remains, yielded 2,560 artifacts, 2,445 of them ceramic. The other middens yielded much more.

Post-Saladoid ceramic vessel from Antigua.
YPM catalog no. 237960.


These ceramic riches have enabled archaeologists to identify 3 successive periods of occupation—roughly pre-A.D. 600, A.D. 600–900, and post-A.D. 900. Each period has its own diagnostic style of pottery. These differences revealed much about the movement of goods and people at different times in the past, the forms of contact and conflict between cultures, and social change within cultures. In particular, the site has offered a unique opportunity to recognize the settlement patterns of 2 overlapping and competing groups in that part of the Caribbean, the Cedrosan Saladoid and Mamoran Troumassoid peoples.

The appearance and eventual disappearance of highly sophisticated Saladoid ceramics at the site tells a fascinating story. Saladoid ceramic-making technology and style, which can be traced back through Trinidad to the Orinoco Valley of what is now Venezuela and nearby Trinidad, penetrated throughout the Caribbean via trade, domination and imitation. But, as the excavation at Indian Creek documents, both the quality of design and the durability of ceramics declined over time. Whether hard times, population pressures, shortage of materials, or simply the loss to memory of necessary skills caused this change, it is eloquent testimony that the passage of time does not always represent progress.

Ceramic bowl rim.
YPM catalog no. 254447.


The evidence uncovered by Rouse and others supports the hypothesis of a Cedrosan Saladoid invasion of the Caribbean from South America. The Cedrosan Saladoid peoples survived in the central and northern part of the Leewards for 300 years after the Mamoran Troumassoid peoples appeared in the Windward Islands and southern Leewards, and after the Ostionoid peoples had emerged in the Virgin Islands and Greater Antilles. They also faced the emergent Elenan Ostionoids across a frontier at the Anegada Passage, as did the Troumassoids who followed. About A.D. 1200 Chican Ostionoid, a newly evolved subseries of cultures in Hispaniola, spread eastward through Puerto Rico and the Virgin Islands to the frontier at Anegada Passage, an expansion that does not appear to have affected the inhabitants of Antigua.

Rouse (upper left) at the Luquillo Sitein Puerto Rico, 1962.


“Culture is something we’ll never know. All we can do is approximate it.
The aim is to learn as much as possible about the past. I didn’t say learn
everything about the past. That would be impossible.” — Ben Rouse

That we know any of this is remarkable. That we know as much as we do results from Ben Rouse’s persistence and patience—and that of his students and colleagues—in the systematic piecing together of Caribbean prehistory, a story he began to investigate over 6 decades ago.